Transnational political violence in African borderlands

Transnational Political Violence in African Borderlands: Expansion Over Relocation

Why do some conflicts stay contained within a single nation, while others bleed across international lines to destabilize entire regions? Understanding this "spillover" is critical for regional security. Researchers studying African conflict patterns have traditionally debated two main ideas. They ask if violence crosses borders because weak states provide "safe havens" (ungoverned areas for rebels) or because strong state counter-insurgency efforts simply push armed groups from one side of a border to the other.

A new study from David G. Russell and Olivier J. Walther proposes a different way to look at this movement. Rather than viewing border crossings as a game of musical chairs where violence merely relocates, the authors suggest that conflicts in Africa tend to grow. Instead of moving from one country to another, they occupy a larger spatial extent on both sides of a border simultaneously.

The limitations of state-centric models

Current interpretations of African political violence often fall into a binary trap. One school of thought argues that "failed states" create sanctuaries that attract transnational terrorists. The opposing view suggests that successful state military actions force these groups to relocate across borders out of necessity.

However, the authors argue that these state-centric explanations fail to account for the active strategies of non-state actors. Rebels and extremist groups are not just passive travelers. They actively compete with the state to control border territories. They also exploit the political marginalization of people living on the periphery. By focusing only on whether a state is "strong" or "weak," previous models have missed the spatial logic of how violence actually grows. As shown in, there is a fundamental distinction between "relocation"—where a conflict moves from point A to point B—and "expansion"—where the conflict occupies a wider area over time.

Mapping the mechanics of conflict

To move beyond these binaries, the authors employ a multi-step spatial analysis. They use disaggregated event data from the Armed Conflict Location & Event Data Project (ACLED). Their methodology relies on two primary analytical pillars:

1. Distance-to-border calculation: The researchers define "borderlands" as any area within 100 km of an international land border. They calculate the precise distance between every recorded violent event and the nearest border. This helps determine if violence clusters near these edges.
2. Spatial Conflict Dynamics indicator (SCDi): This is the core engine of their analysis. The authors overlay a 50km-by-50km grid onto the map. They use the Average Nearest Neighbour Index (ANN)—a statistical tool that measures how much objects cluster or disperse—to classify violence. This allows them to categorize any given area into one of four types based on whether the violence is "clustered" (occurring in tight groups) and "intense" (having a high relative volume).

By combining these tools, the study does not just count events. It characterizes the texture of the violence. This allows the authors to distinguish between a few isolated skirmishes and a sustained, high-intensity insurgency.

Evidence of a distance-decay effect

The findings reveal a stark relationship between geography and instability. The paper reports a clear "distance-decay" effect. This means political violence is significantly more concentrated in borderlands than in the interior. According to the study, the median distance of a violent event to the nearest border is only 99 km. The mean distance is 140 km .

This shows that violence is heavily weighted toward the edges of nations. Roughly 50% of all violent events recorded between 1997 and 2024 occurred within the 100 km border buffer.

The authors also find that borderland violence is qualitatively different. Using the SCDi, the study demonstrates that borderlands are more likely to host "Type 1" violence. This type is defined as being both highly clustered and high-intensity .

In fact, the paper finds that 56% of conflict cells in borderlands fall into this category. This is higher than the 52% found in non-border regions.

Perhaps the most striking result concerns how violence travels. When analyzing the 27 most violent border segments, the authors report 14 clear cases of "diffusion by expansion." They found only 2 cases of "diffusion by relocation" .

This suggests that when a conflict enters a new territory, it does not replace the old one. Instead, it adds to it. This creates a larger, more complex theater of war.

Assessing the spatial blind spots

While the study provides a powerful continental overview, it contains specific limitations. First, the authors base their analysis on "violent events" (such as battles or explosions) rather than fatality counts. They note that this choice may underestimate heavy interstate warfare. For example, in conflicts like the one between Ethiopia and Eritrea, mechanized warfare might cause massive casualties without generating many discrete "events."

Second, the correlation-based method used to distinguish expansion from relocation is a macro-level tool. The authors admit that while this identifies transnational trends, it cannot track the specific tactical movements of individual armed groups. For a commander on the ground, knowing that "expansion is occurring" is helpful. However, it does not reveal which specific militia is moving where.

Finally, the results are highly sensitive to the chosen timeframe. Because conflict geography shifts rapidly, a snapshot of today may not predict the spatial patterns of tomorrow. This is evident in the changing waves of violence seen across different regions .

The verdict: A mandate for integrated security

The evidence suggests that the current trend toward "militarizing" borders is a narrow response. Focusing strictly on restricting movement or conducting airstrikes may not suffice. If conflicts expand rather than relocate, then securing one side of a border is unlikely to stop the growth of the conflict on the other side.

The study's findings imply that security policy must treat borderlands as interconnected, transnational ecosystems. For policymakers, the takeaway is clear. Addressing the root causes of violence requires looking beyond the immediate military threat. Policy should also consider the underlying drivers of marginalization, such as trade disruptions and lack of infrastructure. Until border management integrates developmental support with security capacity, the "expansion" pattern of African conflict is likely to persist.

Figures from the paper